Why should we build up a Start-up vocabulary quickly?

Draft v1.1.  August 4, 2002   Rob Waring

This page is at http://www1.harenet.ne.jp/~waring/vocab/principles/early.htm

Abstract

This article will examine the reasons why it is important both linguistically and psychologically to build a vocabulary quickly when learning a foreign language. The article asserts that very little can be achieved or learned in a foreign language with a small vocabulary and that by building a sizable vocabulary quite quickly one can soon be able to function adequately. You may also wish to look at http://www.jalt-publications.org/tlt/files/95/feb/meara.html

Introduction 

It is obvious that in order to learn a foreign language one needs to learn many many words. But how many? Educated English native speakers have a vocabulary of about 20,000-25,000 word families (A 'word family' refers to a group of words that share the same basic meaning e.g. create, creation, creating, created, creative etc), foreign learners of English need far fewer[1].The speaking vocabulary is usually said to be half of the reading and writing vocabulary. Foreign learners of English only need about 3000-5000 word families to be quite competent in speaking and listening to English. This is great news for learners of English because their task is much easier than that of native speakers!

One of the reasons for this seemingly small number is the nature of words and the frequency with which they appear in a language. Not all words are equal because some words such as time, the, come, make, and so on are very common whereas others such as parasol, bombastic and edifice are relatively rare and not met everyday. It therefore seems clear that these frequent words should be among the first words to learn because they will be met most often and will be needed frequently in speech or writing. Thus the pay off for learning them is higher than for an average rare word. These words are often called a General Service Vocabulary because these words are found in many kinds of situations and domains. This is a vocabulary of about 2000 word families.  The best list (although it is a bit dated is Michael West's 1954 list called the General Service Word List.) These General Service words are found a very very wide range of contexts such as in the medical world, in novels, in scientific reports, on web pages,  in daily conversation, in politics and so on. Because these words appear in so many contexts they are extremely useful to almost all learners. Thus they are called "General Service" words. It has to be remembered of course that each of these topic areas just mentioned has its own specialist or technical vocabulary, for example the words embolism and  gastroenteritis and so on appear in medicine, hydrogen and thermodynamics appear in science and engineering, and interferometer  appears in astronomy.

The case for a General Service vocabulary

Some teachers insist that their learners should only be taught the words that those learners need (e.g. we should teach medical words to doctors) and to teach them words they do not need or do not want to learn is a waste of time.  Quite so. But some of these teachers are not aware that there is a general service vocabulary that is common to ALL the domains. That is, these general service words are the bones on which the meat of the technical vocabulary hangs. Imagine a doctor knowing hundreds of medical words and not knowing get, give, time, break, the and some! Let's be clear about this. From a frequency of input point of view the technical vocabulary does not really start to be a big factor until after the 1500-2000 words of general service vocabulary. In other words in ALL domains the general service vocabulary is more common than the technical words for the first several hundred words at least. Thus teaching doctors only medical words will not allow them to read medical texts because they do not know the General Service words which supports their specialist vocabulary.

Moreover the argument that says that we should teach the words learners needs assumes we know what words they want to learn and assumes we know their individual interests. It is very rare indeed that all the learners in one class are focused entirely and only on one domain and that they will only ever need only the vocabulary from one domain. In any group of learners there will be variations in interest between the learners even if they are ostensibly studying the same domain. This implies that in order to best cover most of the learners, that they all need to learn a general service vocabulary in addition to their specialist interests. Moreover, those learners who want to specialize only in one domain will have to deal with their L2 vocabulary outside their domain in conferences, when traveling, when meeting people and so on.  There is no rationale case against a general service vocabulary for the vast vast majority of learners.

However, it would be misleading to assume that a few hundred words of General Service vocabulary would be enough. This is because there are limits on what one can do with a certain vocabulary. The following schedule lays out a rough picture of what foreign language learners can and cannot do with a certain vocabulary.

So what does all this mean?

It means that without a basic vocabulary there is very very little we can do. We can't understand our environment, and are constantly in need of support from both visual and gestural means in order to have some control over the message.

It also means that one will not be able to do much until one has a basic vocabulary. As one famous person once said "with a limited grammar little, can be said, without words, nothing can be said".

It also means that to reach native-like ability it takes a huge investment of time and of attention to vocabulary.

It appears that early on learners tend to take a word-centred approach to vocabulary acquisition and only later do they move to a more collocational approach (i.e. seeing words with their partners).

Vocabulary is NOT learned once it has been one or two times. We can see that we need to learn the word's multiple uses, the words it goes with and much more. Research also suggests that it takes between 8-20 meetings of a word before we can say that we have 'learned' it.  That means we can understand it when we meet in in our reading and listening but not necessarily use it in our speaking and writing. To use the word productively a lot more knowledge of the word is needed.

There is a difference, a BIG difference, between learning a word only at the meaning-spelling level (or meaning-sound level) and being able to use it.  Thus we need to work extra hard to make the word available productively.

Initially the learners only need a basic picture of the word, such as a rough meaning or a translation will be enough. Later they should work out how the English word is similar and dissimilar from the equivalent word/concept in their own language.

There are certain things we should not be teaching too early. For example, there is little point in teaching affixes until the learner already has mastered the basic root forms.

How have we deal with this situation in the past?

Traditionally many coursebook writers and linguists have stressed the importance of teaching structure.  The argument goes like this. In order for learners to be able to say or write anything they need a framework to put words into.  The framework is grammar. The belief is that learners should first master the grammar patterns and then later add the words. This led to a domination in language teaching of structure with a corresponding de-emphasis on vocabulary. Some coursebooks were marketed on their lack of vocabulary, the intention being that the fewer the words the learners had to deal with, the easier it would be to learn the grammar.  More words were to be added later.

In recent years we have become more and more aware of the fallacy of this approach.  We learned that without vocabulary to put on top of the grammar system the learners could actually say and write very little despite being able to manipulate complex grammatical structures in exercise drills. We also learned that there is no such thing as a broad distinction between grammar and vocabulary. In fact, languages are made up of word patterns (we say weak tea but mild cheese - not weak cheese; we say beautiful day, but not usually handsome day) and chunks of language (by the way, the day after tomorrow, and by and large). We also have hundreds of sentence head (I want to ...., Would you please be so kind as to ..., If I ...., I'll ......, and Have you ever .... ?).  Some of the language is made up of set phrases that never change such as Happy Birthday!; others are semi-fixed where parts can be exchanged such as What ....     like? as it can be rendered as what is he like? , what was the movie like?,  what was the weather like? Research has shown us that a lot of language is formulaic and not really based on the pedagogical grammars that coursebooks teach foreign language learners. Why, for example, do coursebooks teach the present simple tense before the past simple tense when the past simple tense is much more common?

We now also know that our old ideas about words as single items of meaning has been problematic. We used to think that a word was well, just a string of letters with spaces either side and that really was all we needed to worry about. Thus the emphasis in much of teaching has been to teach many meanings. Teachers who say "I taught 10 words today" mean they taught 10 meanings. Certainly there is a meaning-spelling or meaning-pronunciation relationship but there is also a whole lot more than simple knowledge of meanings.  While we have know of other aspects of word knowledge (such as when to use it, its collocations, its frequency of use, etc) we have tended to de-emphasize them in favour of seeing words a simple "teachable" single word items. But with recent research we have discovered the importance of seeing words beyond this level. But meanings exist in multiple word units. For example, Christmas day, How are you?, summer holiday, wet weather, traffic jam, apple pie, sweet and sour pork, and so on.

Thus we are now understand better that there are two basic stages in word knowledge.  The first is the form-meaning relationship (meaning and spelling or sound) and a second "deeper" level of word knowledge that involves knowing how this word the differs from other words; by knowing what contexts it can be used in; whether it is a polite or slang term; by knowing how its meaning changes when used with different word e.g. how a handsome woman (rather large) is different from a handsome man (good looking); and so on. Both of these things need to be worked on, but we cannot deal with the deeper aspects of word knowledge until we have met the word and know its meaning first.

We now have a better understanding that in order for learners to learn a foreign language they need ......

The need for a Start-up vocabulary.

It is clear from the above that learners are not going to get very far without a large vocabulary. Here are just a few reasons why.

Learners expect to learn many words when they start learning a new language. We should take advantage of this expectation

A program of systematic learning will enable learners to see real measurable language gains which will feed a sense of accomplishment and achievement, building confidence.

Learners feel empowered if they have larger vocabularies. They feel they can do things and do not feel so tongue-tied.

Without a start-up vocabulary the teacher (or coursebooks) can't be understood in the foreign language.

Our brains are exceedingly good at finding patterns in language. With a small vocabulary this is hindered because we need to meet many examples of the patterns (particularly tenses and so on) to 'get' them. Finding patterns in 'grammar' will be more difficult at early stages because the learners is not able to meet many examples, thus it is better to build vocabulary first.

Communication strategies will be easier with a larger vocabulary (e.g.  asking for repetition / paraphrasing / restructuring / summarizing / elaboration etc.)

Research evidence shows that we don't start speaking in sentences until a critical threshold of language has been learned (in both 1st and 2nd languages.). Thus the sooner we have those words, the sooner we can leave this stage.

Vocabulary knowledge enables language use which enables vocabulary growth, which in turn enables vocabulary knowledge .............  . etc

Larger vocabularies enable more varied and detailed language experimentation by providing vital feedback on performance. If there is little experimentation / reflection there is little language development

A larger vocabulary allows learners to get to the point where they understand most of a text. Only when about 98-99% (1 new word in 100 or 1 in 50) of a text is understood can one consistently and successfully guess new word meanings from the context. Research shows that below this level, successful guessing without the need to turn to a dictionary is extremely difficult. Thus in order to be able to start reading this basic vocabulary, the words should be in place already.

Learners avoid aspects of language use and avoid errors due to a lack of vocabulary. Typically beginner foreign language learners talk using pronouns and nouns and tend to avoid verbs or use only base (uninflected) verbforms. In other words they speak in content words and do not inflect them for time, aspect or case because they are focused on getting meanings across. Attention to "correcting their own output" comes later.

We cannot say much in class or to our foreign friends unless we have a few words to do it with. Building a Start-up vocabulary will allow the learner to manage their own learning

The more vocabulary the learners have the earlier in their career, the sooner they will be able to see words as chunks, or formula instead of as individual words. As has been discussed, much of language is not "grammar v's vocabulary" but is made of chunks of language which exist in many forms. 

So, what is a Start-up vocabulary?

A Start-up vocabulary would be a basic set of words that the learner needs in order to start to learn from books or by listening to the teacher. Learners with little or no vocabulary can do very little in the target language and as soon as they get a few hundred basic words, then they will be able to learn from their coursebooks and understand their teacher in the L2. As a complete beginner, we cannot of course just pick up a book and try to read from that as we do not know the spelling-sound combinations.

A Start-up vocabulary is those words and phrases which one needs in order to manage one's own learning.  This might include

say 500-1000 basic word families (there are some lists here)

a few set phrases such as "What does ...mean?" , 'How do you spell ...?' and 'How are you?' or  'Nice to meet you' and so on

some words for understanding their classes and text books such as noun  verb chalk, board homework etc.

some words from their textbooks that they will be using at the start of their course.

How can we learn this startup vocabulary?

There are some people who suggest that we can build the vocabulary fairly quickly by reading. Most of the evidence for this notion comes from experiments that have looked at the speed at which children learning their mother tongue pick up words from reading. They typically find that children learn on average about 1000 new word families per year. So by the time they are 20 they already have a sizable vocabulary.

However, we have to be very clear that learning to read in the mother tongue and learning to read in a foreign language is very different. Firstly, by the time a child is ready to learn to read in her mother tongue, she already knows thousands of words and reading is a mostly matter of matching already known sound labels with written labels. As they already have a huge vocabulary resource of several thousand words to call upon to build a context for the unknown words, guessing new words is quite easy. By contrast, foreign language learners do not have the benefit of a huge listening or reading vocabulary to start reading with. This vocabulary needs to be built from scratch.  Secondly, in the same way the foreign language learner does not already have a grammatical system in place to hang new words on, this also has to be learned. Thirdly, children do not learn in the same ways that adult foreign language learners do.  Adult language learners have the benefit of analytical skills that children do not posses, but they also have lost the ability to learn freely and easily like children do. Therefore, it is probably more effective to teach the basic vocabulary first to enable learning from the reading (or listening) to take place.

An alternative to learning from reading is to learn from lists of words. This vocabulary can be build fairly rapidly if there is a systematic approach to learning these words. (This is discussed in detail in another article.) However, in brief, the article discuses a system for memorizing many words very quickly and sets out some basic principles of memory that can aid this learning. Research clearly shows that people learn words much faster out of their context than in it. That is, words that are in lists or on word card (flashcards) (this is called direct learning, discrete item learning or decontextualized learning) can be learned faster than from reading them in books. Studies dating back 100 years have shown a consistent advantage for this type of learning.

Fitting a Start-up vocabulary into a curriculum.

In order to build to vocabulary the learner needs to be introduced to the task at hand. These initial tasks include

finding the words they need

memorize them

practice them in different and meaningful ways

learning how to learn the words effectively

Below is a suggested overview of how the discrete item learning of words can be done in tandem with other things.  The following diagram is split horizontally into "in class" activities and "out of class" activities.

Initially in this schedule there is a very heavy emphasis on vocabulary learning and if the learning is done effectively this need not last more than a few weeks. Also important in this process is to see an early emphasis on building a study atmosphere, explaining why it is necessary to learn a lot of words quickly so they can quickly start to read and listen later, and explaining how to do it. Most of this learning is done out of class, and there would need to be constant need to ensure that the learning was being done effectively.  Learners can test each other and competitive classes may have a reward / competitive element if desired.  The aim at this stage is to build the Start-up vocabulary, so that learners can then move on to other things.

As the ability level of the learner increases, different activities come to dominate. There will be more normal classwork which the learners will now be able to tackle faster and easier, and a LOT of work on simplified reading and listening to build fluency with the words learned and to consolidate them. The following table sets out the steps necessary to master the types of vocabulary one will need.

Note 1: a Wordbank is an envelope in which the learner keeps her own blank pieces of paper. She writes new words on one side of the card and puts the meaning on the back.  Before class starts (or at some other time) she should go through them to remember them. It is like saving money in a bank i.e. building for the future.

Note 2: It is not a good idea to learn word parts (affixes, Latin and Greek roots and so on) too early.  It can be tempting as there appear very generative (i.e the ideas is that if you learn the system then you can easily find the hints to words).  Research shows that only advanced learners have sufficient background to take advantage of this (even if their language very closely resembles English).

[1] There has been quite a lot of disagreement over the years as to the size of an English native speaker's vocabulary, some say it is 60-70,000 words and others as high as 120,000, but these figures are derived from the different ways that words are counted. For example, some people count every word form as separate (thus help helps and unhelpful would all be considered separate words and therefore the number of words one knows will be much higher). Other problems with counting words come from how one counts words in dictionaries. Usually a vocabulary size is counted by looking at the percentage of words known in a huge non-technical dictionary, but unfortunately there are many problems and pitfalls often overlooked in doing so. One person who found a very large vocabulary  derived the figures without counting words in the dictionary by believing the publisher's assertion of the number of words, when later it was found (using a better definition of what we count as a word), that the publisher had greatly inflated the number, presumably for advertising reasons).